Journal of medical ethics 1994; 20: 146-151

First trimester prenatal diagnosis: earlier isnot necessarily betterJudith A Boss University ofRhode Island, USA

Author's abstractIn the past few years considerable attention has beengiven to a relatively new method ofprenatal diagnosisknown as chorionic villus sampling (CVS). BecauseCVS can be performed in the first trimester it is hailedby many as a significant advance over amniocentesis.What has not been as publicized, however, are thedisadvantages ofCVS and earlier prenatal diagnosis.The emotional costs ofCVS in terms of the greaternumber of both spontaneous and selective abortionsfollowing CVS, the use ofCVS for sex selection and,because of the greater social acceptability offirsttrimester abortion, the possibility of increased pressure onwomen to undergo prenatal diagnosis by healthinsurance companies, medical professionals andgovernment agencies, all need to be weighed against theadvantages of early prenatal diagnosis.

Second trimester amniocentesis is presently the mostwidely used method of prenatal diagnosis. In thepast few years considerable attention has been givento a relatively new method of prenatal diagnosisknown as chorionic villus sampling (CVS).

Unlike amniocentesis, which cannot be performeduntil the second trimester of pregnancy or very latein the first trimester, CVS can be carried out as earlyas eight to ten weeks' gestation. Also, while amnioticfluid needs to be cultured for three to four weeks,analysis of the chorionic villus sample can be carriedout in twenty-four hours.

Because of this many medical professionals haveenthusiastically endorsed the use of CVS overamniocentesis. 'It would be expected', concludesone group of prominent physicians and researchers:

'that the decision to interrupt the pregnancy in thesecond trimester would be more difficult than afterCVS... . First trimester diagnosis allows privacy inreproductive decisions as the pregnancy is not yetphysically evident and announcements of the preg-nancy to family and friends may be delayed untilprenatal evaluation is completed. Moreover, first

Key wordsAbortion; amniocentesis; CVS; prenatal diagnosis.

trimester termination of pregnancy is safer andprobably less damaging emotionally than secondtrimester termination' (1).

These advantages have led to a significant shift inthe past few years towards using CVS (2,3). Whathas not been as publicized, however, are thedisadvantages of CVS and first trimester prenataldiagnosis. It is the purpose of this paper to point outand discuss some of these possible disadvantages.

AccuracyOne of the drawbacks of CVS is its lower accuracyrate compared to the better than 99.5 per cent ratefor amniocentesis (4). This difference cannot beexpected to improve substantially since CVS usescells from the placenta rather than directly utilizingfetal cells for analysis as does amniocentesis. About3.5 per cent of CVS procedures need to be repeatedor followed up by amniocentesis because of failure toobtain adequate material for analysis, matemal cellcontamination, and/or mosaicism - a conditionwhere chromosomal abnormalities only appear insome of the cells (5,6,7,8).

Fetal lossAlthough CVS appears to pose little risk to thewoman, there is a 3.2 per cent procedure-relatedfetal loss; with amniocentesis this rate is less than 1per cent (4,5,6,7,9). Fetal loss appears to be highestwith transcervical CVS (10,11). Evans andcolleagues estimate that the procedure-relatedcomplication rate, in terms of fetal loss, is almost 5 /2times higher for CVS than for amniocentesis, a riskwhich they find 'acceptable' (3).

By undergoing CVS a woman accepts a one inthirty chance of miscarrying what is probably anormal fetus (3). In other words, if the difference inrisk between CVS and amniocentesis is 2 per cent,an estimate which is probably on the low side, therewill be as many as 66 normal fetuses lost for every100 abnormalities detected ( 11). Initially it wasconjectured that the procedure-related loss woulddrop as operators became more proficient in

by copyright. on S

eptember 28, 2020 by guest. P

rotectedhttp://jm

e.bmj.com

/J M

ed Ethics: first published as 10.1136/jm

e.20.3.146 on 1 Septem

ber 1994. Dow

nloaded from

J7udith A Boss 147

performing the procedure (6). However, this has notproved to be the case (9).

Fetal damageAnother disadvantage ofCVS is a possible connectionbetween CVS and malformations ranging from minorones such as strawberry hemangiomas - reddishdiscoloration of areas of the skin - to majormalformations such as limb reduction (2,4,12). In a

recent study of the pregnancy outcomes of 436patients who underwent CVS, there were 391surviving infants, 18 selective abortions,* and 27 fetalor neonatal deaths (10). Twenty-three of the lastgroup involved chromosomally normal fetuses. Ofthe394 fetuses and infants who were evaluated 3.3 per

cent had major congenital anomalies, including fourcases of limb reduction. All four cases of limbreduction had normal karyotypes.A similar study found five cases of limb reduction

among 289 pregnancies in which CVS had beencarried out during the first trimester (13). Theserates appear to be well above what would normallybe expected.

Higher rate of 'unnecessary' selectiveabortionsThe earlier stage at which CVS is performedsignificantly increases the chances of finding a fetalabnormality. The rate of spontaneous fetal loss afterclinical diagnosis of pregnancy is 12-15 per cent.Most of this loss is the result of chromosomalabnormalities in the young fetus or embryo (14). Forexample, more than 99 per cent of conceptions withTurner syndrome end in spontaneous abortion (15).And an estimated two-thirds of all fetuses withDown syndrome spontaneously abort during thefirst trimester (16).

After sixteen weeks, however, the rate ofspontaneous loss is only 1 per cent. A test performedat eight weeks' gestation, consequently, will findover five times as many chromosomal abnormalitiesas a test done at sixteen weeks (17). This means thatfive times as many parents will have to suffer throughthe anguish of having to decide whether to aborttheir fetus. Thus, while the maternal mortality rateof second trimester abortions is higher (18), thisfactor is offset by the higher number of abortionsthat would be performed on women undergoing firsttrimester CVS.

Psychological trauma of selectiveabortionProspective parents are rarely prepared by thegenetic counsellor for the extent of the psychologicaltrauma experienced after a selective abortion. Whilethere is no intention to minimize the tragedy andgrief associated with losing a fetus through

spontaneous abortion, the psychological effects of amiscarriage on parents are not as severe or long-lasting as those of a selective abortion (19).

Because first trimester selective abortion is arelatively new phenomenon there have not, as yet,been any controlled studies to test the hypothesis thatfirst trimester selective abortion is 'probably lessdamaging' (1) than an abortion performed later in thepregnancy. One cannot legitimately generalize fromfirst trimester elective abortions, where the pregnancyitself is unwanted, to first trimester selective abortionswhere a wanted pregnancy is terminated.

Despite the widespread belief that the merepresence of 'choice' in the decision to have anabortion significantly reduces the anguish felt by thefamily, those who choose abortion because the fetushas a congenital disorder often suffer long-term griefwhich is almost identical to that experienced over thedeath of a newborn infant (19,20,21). Unlike mostmiscarriages or stillbirths, where the loss is passive,selective abortion requires a deliberate choice. Amiscarriage following CVS may also be perceived bythe mother as her doing or 'fault', rather than as amiscarriage due to natural causes, because she choseto expose the fetus to risk by undergoing CVS. Theethical conflict of having to choose against life orfor suffering, and the realization of one's owncontribution to the ending of a life can lead to a lossof moral self-esteem and feelings of failure and guilt(22).Another important difference between a stillbirth

or death of a born child and the loss of a fetus isthat the grief over the death of an aborted fetus isgenerally not recognized by society. Therefore, it isoften harder for the parents to work through theirfeelings of loss following an abortion (23). It ispossible that the 'privacy' of first trimester prenataldiagnosis and selective abortion may actuallyincrease the unresolved 'disenfranchised' griefsince so few people will know about the person'sloss.A recent study of the psychosocial sequelae of

second trimester selective abortions found that theyrepresented 'an emotionally traumatic major lifeevent for both mother and father' (22). Even aftertwo years 20 per cent of the women surveyed 'stillcomplained of regular bouts of crying, sadness andirritability'. Twenty-eight per cent of the parents alsowondered if the physician had made a wrongdiagnosis. This percentage might be even higheramong a group who had had an abortion followingCVS, since the accuracy rate of CVS is lower thanthat of amniocentesis.

*The term 'selective abortion' is being used here to refer toan abortion where the parent(s) choose to terminate awanted pregnancy because of a genetic disorder or someother undesirable characteristic of the fetus. 'Electiveabortion' refers to the planned termination of an unwantedpregnancy.

by copyright. on S

eptember 28, 2020 by guest. P

rotectedhttp://jm

e.bmj.com

/J M

ed Ethics: first published as 10.1136/jm

e.20.3.146 on 1 Septem

ber 1994. Dow

nloaded from

148 First trimester prenatal diagnosis: earlier is not necessarily better

The emotional costs of CVS, therefore, must beweighed against the benefits of early diagnosis inlight of the fact that the great majority of fetusesdiagnosed with genetic disorders would havespontaneously aborted in the next month or twoanyway. Because spontaneous abortion is lesstraumatic than a first trimester selective abortion,Hecklering and Verp calculate that second trimesterprenatal diagnosis might be preferable in all butthose cases where the fetus has at least a 50 per centrisk of carrying a genetic disorder (24).

The fetal therapy justificationOne justification used for early prenatal diagnosis isthat it allows physicians to diagnose and treat fetaldisorders. There are efforts being made, it is pointedout, to develop in utero therapeutic procedures forcorrecting fetal abnormalities. One of the mostpromising is fetal stem-cell transplantation whichinvolves grafting cells early in pregnancy while thefetus is still immunologically tolerant and before thedisorder has become too developed (25). Also,with the improvement of high-resolution, real-timeultrasound, the success rate of intravascular bloodtransfusion for certain haemolytic (blood) disordersis now 85 per cent to 95 per cent (26).The fetal therapy justification is very appealing to

parents. In one study 185 women were asked to ratetheir attitude towards different types of prenataldiagnosis. The most favourable rating was givenfor 'ultrasound examination aimed at detectingtreatable abnormalities' (27).

There is little doubt that relatively non-invasivetechnology whose primary purpose is to diagnosistreatable disorders would be warmly welcomed byparents and ethicists alike. However, this is not thereality associated with CVS. While new develop-ments in fetal therapy hold out future hope forfetuses with disabilities, the present reality is thatprenatal diagnosis rarely leads to fetal therapy. Infact, far more normal fetuses are killed as a result ofthe CVS procedure.

Even if therapy were possible, the availability ofearly prenatal diagnosis may make it easier and lesscostly simply to 'privately' abort the abnormal fetusand 'start again', rather than go through the greatertrouble, uncertainty and expense of in uterosurgery. Medical practitioners and pharmaceuticalcompanies may also become less motivated todevelop treatments for conditions that don't have tooccur (28).

Research on cures for genetic disorders, whetherin utero or after birth, may also be put aside in favourof selective abortion. This has already startedhappening to some extent. During the 1960s, forexample, there were two to three times as manypeople working on a cure for Tay-Sachs disease thanat present. The emphasis now is put on a prenataldiagnosis for Tay-Sachs disease, followed by

abortion in the case of a positive diagnosis (29).Similarly, as soon as a prenatal diagnostic test forHuntington disease became available in the early1 980s, 'funds began to disappear for research to finda cure' (30). This trend might accelerate now thatfirst trimester prenatal diagnosis is available.Coupled with a decreasing amount of public moneysbeing used to assist the disabled, this could makecontinuing a pregnancy with a fetus who has adisabling disorder an option open only to moreaffluent parents.

It is generally conceded by the medical professionthat the primary aim of prenatal diagnosis is thedetection, and subsequent abortion, of abnormalfetuses. Because of the procedural risks to the fetusand the lack of effective methods of fetal therapy, '...with regard to most malformations, prenataldiagnosis is a rational activity only if abortion is seenas an acceptable alternative' (31).

The 'reassurance' justificationA more commonly used justification of earlyprenatal diagnosis is that it provides a means ofreassuring anxious parents that their fetus is normal.However, while CVS might reassure parents thattheir child does not have certain chromosomal andgenetic disorders, one disadvantage of CVS is that,unlike amniocentesis, it cannot be used to diagnoseneural tube defects (NTDs) such as spina bifida andanencephaly. Although the majority of debilitatingcongenital disorders are the result of chromosomaland genetic anomalies, most of which can bedetected by CVS, the incidence rate for NTDs is stillsignificant at between 1.0 and 1.6 per 1,000 births -about the same as for Down syndrome, the mostcommon major chromosomal disorder in newborns(32,33). Also, unlike genetic disorders which aregenerally inherited, 95 per cent of all infants withNTDs are born to parents with no previous familyhistory of the disorder (32).

Consequently, even if parents do have a negativetest result from CVS and have no family history ofNTDs they cannot rest assured that their fetus doesnot have an NTD until the second trimester whenthe alpha-fetoprotein (AFP) level can be tested usingeither amniotic fluid or maternal serum. Even thenAFP testing is only a screening tool rather than adiagnostic tool. Ultrasound, which can pick up mostcases of neural tube defects, is also not an effectivediagnostic tool prior to 12 weeks (34,35).

Barbara Rothman, in her book The TentativePregnancy, maintains that it is the medical professionthat has created a need for prenatal diagnosis for'reassurance' by creating what she calls 'geneticanxiety', thereby capitalizing on women's normalfear of having a 'defective', socially unacceptablechild - just as deodorant and mouthwash companiesfirst had to create anxiety about socially unaccept-able body odour before they could market their

by copyright. on S

eptember 28, 2020 by guest. P

rotectedhttp://jm

e.bmj.com

/J M

ed Ethics: first published as 10.1136/jm

e.20.3.146 on 1 Septem

ber 1994. Dow

nloaded from

J7udith A Boss 149

product (29). One cannot help but wonder if thebattery of tests pregnant women are 'encouraged' tosubmit to for their own 'peace of mind' are notmore to reassure the physicians that they will not beheld legally liable should a 'defective' fetus get byundetected (36) and, in the case of first trimesterCVS, that they will not have to perform lateabortions and will thereby avoid the often greatpsychological trauma on the medical staff who haveto perform late abortions (37,38).

Abortion for sex selection and minordisordersThe greater social acceptability and 'privacy' of firsttrimester abortion may also lead to an increase ofabortions for sex selection and for minor defects. Inour society, where discrimination is based not onlyon physical and mental abilities, but on gender aswell, female fetuses have a lower value (28). In astudy of 2,278 women who had had CVS, it wasfound that CVS was being used more often thanamniocentesis for reasons such as sex selection,despite policies against the use of prenatal diagnosisfor sex selection at the centres which participated inthat study (1).

In another study comparing CVS and amnio-centesis, eight ofthe 1,928 women who had CVS laterhad an abortion for 'social reasons' (11). These eightmade up 19 per cent of the total number of selectiveabortions. Of the 959 women who had undergoneamniocentesis, none of the women had selectiveabortions for reasons other than chromosomalaberrations and sonographic abnormalities.

While approval of abortion for sex selection is lowin the United States and most Western countries,the declining size of the American family, thepreference for sons as first born or only children, andthe emphasis on reproductive autonomy, are allcontributing to a modification of this attitude. A1975 survey of genetic counsellors in the UnitedStates found that, except in the case of sex-linkedgenetic disorders such as haemophilia or Duchennemuscular dystrophy, only 1 per cent would performprenatal diagnosis for sex selection (37). By 1988this figure had climbed to 62 per cent with themajority of geneticists giving as their primary reason'respect for women's autonomy' (38).The practice of withholding information about

the gender of the fetus, unless it is medicallyrelevant, is viewed by some as patronizing and aninfringement on the parents' reproductive freedomof choice - a choice which should include deciding ifthe gender of the fetus constitutes too great a socialburden for the parents. Other geneticists andfeminists, on the other hand, argue that the use ofprenatal diagnosis for sex selection is the 'originalsexist sin ... because it makes the most basic moraljudgement about the worth of a human being restfirst and foremost on its sex' (39).

While many people claim that abortion choicesare private rather than moral choices, it is hardto reconcile the medical profession's traditionalcommitment to egalitarian principles with the viewthat female feticide is a value-neutral choice. Thefamily is not a private autonomous unit but rather agroup existing within a particular social context, acontext where being female constitutes an inferiorstatus. If a woman is under pressure, whetherreligious or cultural, to have a son, should physiciansyield to this pressure, thereby legitimating existingsex discrimination, or do they have a moralobligation to work for a more egalitarian society?The principle of egalitarianism also raises the

question of whether the worth of a human beingshould rest solely on her or his physical or mentalabilities, especially where these disabilities are not life-threatening or incompatible with sentient life. Womenwho have CVS seem to be more willing to seek anabortion for relatively minor disorders. For example,one study found that 97-6 per cent ofwomen who hada positive result from CVS chose abortion, while only78- 1 per cent of those who had a positive result fromamniocentesis opted for abortion (40).The privacy of first trimester prenatal diagnosis

and the relative ease of a first trimester abortion mayeventually lead to demands for a 'custom-madechild' where abortions are sought for fetuses whomight have a genetic tendency towards obesity, or'only' average intelligence and athletic prowess (41).'The allure of a genetic test for a normal, or, in thefuture, an optimal baby', geneticist Marc Lappecautioned us in 1973 when prenatal diagnosis wasstill in its infancy, 'threatens to reinforce aninexorable trend in Western society towardstypecasting the less-than-optimal into categories forassortment and ultimate disposal' (42).

Pressures to undergo prenatal diagnosisThe availability of first trimester prenatal diagnosismay also increase the social pressure on women, whowould otherwise prefer not to undergo the invasivemedical procedure, to seek prenatal diagnosis. In arecent survey of 185 women who had just given birthto their first child, 30 per cent replied that theywould rather not have CVS or amniocentesis duringtheir second pregnancy. Only 36 per cent said theywould definitely have one or the other during theirsecond pregnancy (27).

'The social pressures to participate in screeningprograms and to terminate an affected pregnancy areconsiderable', notes child care and developmentspecialist, Dr Josephine Green. 'Women may, as aresult, experience considerable stress and find them-selves acting against their own moral convictions'(43). Since the rate of chromosomal anomaliesincreases dramatically after age 30-35, so-called'older women' would be particularly vulnerable tosuch pressures. The introduction of first trimester

by copyright. on S

eptember 28, 2020 by guest. P

rotectedhttp://jm

e.bmj.com

/J M

ed Ethics: first published as 10.1136/jm

e.20.3.146 on 1 Septem

ber 1994. Dow

nloaded from

150 First trimester prenatal diagnosis: earlier is not necessarily better

biochemical screening using maternal serummarkers for disorders such as Down syndrome,while presently in the experimental stages (44,45),could bring additional pressure to bear on women toundergo screening and, should the screening resultsbe positive, to seek early prenatal diagnosis andpossibly abortion.The pressures brought to bear on women by the

very availability of first trimester prenatal diagnosismay be similar to those created by screeningprogrammes for such disorders as Tay-Sachs andsickle cell anaemia. Although these programmes wereintroduced to help parents in their family planning,the very availability of these programmes has led topressure on those in high-risk categories to undergoscreening, as well as to the stigmatization of carriersand social disapproval of marriage between carriers.

Bioethicist John Fletcher fears that, because of thecost-benefit advantage of prenatal screening, 'aclimate of moral blame' may be cast around parentswho do not comply. Indeed, suggestions havealready been made that legislation should be passedto prevent the birth of children who fall below a'minimum standard' (47). Because of the greatersocial acceptability of first trimester abortions,health insurance companies may also feel justified inwithholding medical coverage from women whocould have avoided the birth of a child with a 'costly'genetic disorder (48,49).Some physicians and ethicists express concern

that justifying selective abortion of fetuses withgenetic disorders on the basis of their potentialfinancial or social burden to society may eventuallylead to a policy of mandatory genetic screening andprenatal diagnosis (46,50). Public approval ofabortion for genetic disorders is significantly higherif the abortion is carried out during the first, ratherthan the second, trimester (51). This, in conjunctionwith the strong cultural bias in the Western worldtowards technological control of our environmentand destiny, may make 'quality control' of ouroffspring not simply an option, but an obligation.The state of California already requires that all

pregnant women be 'offered' alpha-fetoprotein(AFP) testing. Those women who do not want AFPtesting have to sign a refusal form (30). Should firsttrimester CVS become widely accepted similarrequirements could be made regarding itsavailability to all pregnant women.

Incorporating prenatal diagnosis, and the issue ofabortion, on the path to motherhood further removescontrol over one's pregnancy from women and placesit even more in the hands of the medical professionand technocrats (29). Harvard biologist RuthHubbard writes in this regard that 'a woman whodecides not to have prenatal tests or not to abort afetus whom she knows has a disability takes on respon-sibility for the social, medical, and economic problemsshe and her family may experience as a result. It isincongruous to call that making a choice' (30).

ConclusionIn a consumer-oriented society that values qualityproducts, quick fixes and 30-minute solutions tomajor life crises, we must not forget to step back andre-examine the 'the sooner the better' assumption.The advantages of first trimester CVS must beweighed against the cost of the greater number ofprocedure-related miscarriages and often unneces-sary abortions that will be performed. Even forcouples who are both carriers of deleteriousautosomal recessive genes or, in the case of thewomen, an X-linked disorder, the choice of earlierprenatal diagnosis is not as obvious as might be firstthought. The cost of losing or harming a normal, verymuch wanted child as a result of CVS must beweighed against the burden of the extra two monthsofpregnancy before amniocentesis can be performed.

In addition, the very advantages in terms ofprivacy and the social acceptability of first trimesterabortions may not only encourage the use of prenataldiagnosis for minor disorders and non-medicalreasons, such as sex selection, but could actuallylimit a woman's options by leading to policies thatput pressure on women to undergo prenataldiagnosis and abortion of fetuses with defects.Earlier is not necessarily better.

JYudith A Boss, MA, MSc, PhD, is a PhilosophyLecturer in the Philosophy Department of the UniversityofRhode Island, Kingston, USA.

References(1) Rhoads G, Jackson L G, Schlesselman S et al. The

safety and efficacy of chorionic villus sampling forearly prenatal diagnosis of cytogenetic abnormalities.New England journal of medicine 1989; 320: 609-617.

(2) Johnson A, Wapner R J, Davis G H, Jackson L G.Mosaicism in chorionic villus sampling: an associ-ation with poor perinatal outcome. Obstetrics andgynecology 1990; 75: 573.

(3) Evans M I, Drugan A, Koppitch F C, Zador I E,Sacks A J, Sokol R J. Genetic diagnosis in the firsttrimester: the norm for the 1 990s. American journal ofobstetrics and gynecology 1989; 160: 1332-1339.

(4) Medical Research Council. European trial of chorionicvillus sampling. Lancet 1991; 337: 1491-1499.

(5) Canadian Collaborative CVS-Amniocentesis ClinicalTrial Group. Multirandomised clinical trial ofchorion villus sampling and amniocentesis. Lancet1989; 1:1-6.

(6) Bergman B, Claesson U, Mark J. Risk evaluation in asmall series of transabdominal chorionic villussampling. Gynecologic and obstetric investigations 1991;31: 189-191.

(7) Mitchell G, Loughman W, Epstein C, Halbasch G,Stephens J, Hall B. Prenatal genetic diagnosis in3,000 amniocenteses. New Englandjournal of medicine1979; 300: 157-163.

(8) Miny P, Hammer P, Gerlach B, Tercanli S, Horst J,Holzgreve W. Mosaicism and accuracy of prenataldiagnosis after chorionic villus sampling and placentalbiopsies. Prenatal diagnosis 1991; 1 1: 581-589.

by copyright. on S

eptember 28, 2020 by guest. P

rotectedhttp://jm

e.bmj.com

/J M

ed Ethics: first published as 10.1136/jm

e.20.3.146 on 1 Septem

ber 1994. Dow

nloaded from

Judith A Boss 151

(9) Meade T W, Grant A M. Chorionic villus sampling.British medicaljournal 1992; 304: 185-186.

(10) Burton B K, Schulz C J, Burd L I. Limb anomaliesassociated with chorionic villus sampling. Obstetricsand gynecology 1992; 79: 726-730.

(11) Smidt-Jensen S, Philip J. Comparison oftransabdominal and transcervical CVS and amnio-centesis: sampling success and risk. Prenatal diagnosis1991; 11: 529-537.

(12) Kaplan P, Normandin J J, Wilson G N, Plauch H,Lippman A, Vekemans M. Malformations and minoranomalies in children whose mothers had prenataldiagnosis: comparison between CVS and amnio-centesis. American journal of medical genetics 1990; 37:366-370.

(13) Firth HV, Boyd PA, Chamberlain P, MacKenzie I Z,Lindenbaum H, Huson S M. Severe limb abnormali-ties after chorion villus sampling at 56-66 days'gestation. Lancet 1991; 337: 762-763.

(14) Boue J, Boue A, Lazar P. Retrospective andprospective epidemiological studies of 1,500 karyo-typed spontaneous human abortions. Teratology1975; 12:11.

(15) Rogers D, Shapiro L. X-linked disease and disordersof the sex chromosomes. In: Milunsky A, ed. Geneticdisorders and the fetus. New York: Plenum Press, 1986:349.

(16) Creasy M R, Crolla JA. The management of thepatient having fetal diagnosis. Clinical obstetricgynaecology 1974; 1: 737-745.

(17) Simpson J L. Incident and timing ofpregnancy losses:relevance to evaluating safety of early prenataldiagnosis. American journal of medical genetics 1990;35: 165-173.

(18) Tyler C. Epidemiology of abortion. Jrournal ofreproductive medicine 1981; 26: 463.

(19) Blumberg B, Golbus M, Hanson K. Thepsychological sequelae of abortion performed for agenetic indication. American journal of obstetrics andgynecology 1975; 122: 799-808.

(20) Powledge T. Toward a moral policy for sex selection.In: Bennett N, ed. Sex selection of children. New York:Academic Press, 1983: 40.

(21) Borg S, Lasker J. When pregnancy fails: coping withmiscarriage, stillbirth and infant death. Boston: BeaconPress, 1981.

(22) White-Van Mourik M C S, Connor J M, Ferguson-Smith M A. The psychological sequelae of a secondtrimester termination of pregnancy for fetal abnor-mality. Prenatal diagnosis 1992; 12: 189-204.

(23) de Spelder L, Strickland A. The last dance. Palo Alto,CA: Mayfield Publishing Co, 1987: 218-219.

(24) Heckerling P S, Verp M S. Genetic risk and earlyversus late prenatal testing. Lancet 1991; 337: 363.

(25) Ferguson-Smith M A. Early prenatal diagnosis.Prenatal diagnosis 1991; 11: 489-491.

(26) Grannum P A, Copel J A. Invasive fetal procedures.Radiologic clinics ofNorth America 1990; 28: 217-226.

(27) Tymstra T J, Bajema C, Beekhuis J R, Mantingh A.Women's opinions on the offer and use of prenataldiagnosis. Prenatal diagnosis 1991; 11: 893-898.

(28) Lippman A. Prenatal genetic testing and screening:constructing needs and reinforcing inequities.American journal oflaw and medicine 1991; 17: 15-50.

(29) Rothman B. The tentative pregnancy. New York:Viking Press, 1986: 230.

(30) Hubbard R. Prenatal diagnosis and eugenic ideology.Women's studies internationalforum 1985; 8: 575.

(31) Carmenisch P. Abortion: for the fetus's own sake?Hastings Center report 1976; 6: 38.

(32) Milunksy A. The prenatal diagnosis of neural tube andother congenital defects. See reference (15): 453-520.

(33) Hsu L Y F. Prenatal diagnosis of chromosomeabnormalities. See reference (15): 115-183.

(34) Romero R, Pilu G, Jeanty P, Ghidini A, Hobbins J.Prenatal diagnosis of congenital anomalies. Norwalk,CT: Appleton and Lange, 1988: 45.

(35) Nicolaides K H, Campbell S. Diagnosis of fetal abnor-malities by ultrasound. See reference (15): 521-570.

(36) Furhman R, Black R. Pregnancy termination forgenetic indications: the impact on families. Socialwork in health care 1984; 10: 17-34.

(37) Fletcher J, Wertz D. Ethical aspects of prenataldiagnosis: views of US medical institutes of health.Clinicalperinatology 1987; 30: 30.

(38) Wertz D, Fletcher J. Ethics and medical genetics inthe United States. American journal of medical genetics1988; 29: 323.

(39) Powledge T. Unnatural selection: on choosingchildren's sex. In: Holmes H, Hoskins B, Gross M,eds. The custom-made child? A woman centered perspec-tive. Clifton, NY: Humana Press, 1981: 193-199.

(40) Verp M S, Bombard A T, Simpson J L, Elias S.Prenatal decision following prenatal diagnosis of fetalchromosome anomalies. American journal of medicalgenetics 1988; 29: 613-622.

(41) Holmes H, Hoskins B, Gross M. The custom-madechild? A woman centered perspective. Clifton, NJ:Humana Press, 1981.

(42) Lappe M. How much do we want to know about theunborn? Hastings Center report 1973; 3: 8.

(43) Green J M. Prenatal screening and diagnosis: somepsychological and social issues. British journal ofobstetrics and gynaecology 1991; 97: 1076.

(44) Johnson A, Cowchock F S, Darby M, Wapner D,Jackson L. First trimester maternal serum alphafeto-protein and chorionic gonadotropin in aneuploidpregnancies. Prenatal diagnosis 1991; 11: 443-450.

(45) Mancini G, Perona M, Dall'Amico D et al. Screeningfor Down's syndrome with maternal serum markers -an experience in Italy. Prenatal diagnosis 1991; 11:245-252.

(46) Fletcher J. Ethics and trends in applied humangenetics. Birth defects 1983; 19: 143-158.

(47) Fletcher J. Our duty to the unborn. In: Fletcher J, ed.Humanhood: essays in biomedical ethics. Buffalo, NY:Prometheus Books, 1979: 123-124.

(48) Gostin L. Genetic discrimination: the use ofgenetically based diagnostic and prognostic tests byemployers and insurers. American journal of law andmedicine 1991; 17: 109.

(49) Baum R. Genetic screening: medical promise amidlegal and ethical questions. Chemistry and engineeringnews 1989 Aug 7: 10-16.

(50) Johnson J, Elkins S. Prenatal treatment: medical andgene therapy in the fetus. Clinical obstetrics andgynecology 1988; 31: 409.

(51) Lamanna M A. Social science and ethical issues: thepolicy implications of poll data on abortion. In:Callahan S, Callahan D, eds. Abortion: understandingthe differences. Hastings-on-the-Hudson, NY: 'TheHastings Center, 1984: 4-5.

by copyright. on S

eptember 28, 2020 by guest. P

rotectedhttp://jm

e.bmj.com

/J M

ed Ethics: first published as 10.1136/jm

e.20.3.146 on 1 Septem

ber 1994. Dow

nloaded from